Featured
A design review of a Google Cloud image pipeline: signed uploads, fixed variants, immutable names and a CDN, with failure and deletion paths worked through.
15 min
Topic 01 / 08 · 09 articles
About this topic
152 words
Most cloud architecture advice is written for companies with a platform team and a generous budget. The articles here are written for the opposite situation: a small team, a real product, and a bill that has to make sense every month.
I focus on Google Cloud because it is where I run production systems, but the reasoning travels. Each piece starts from a concrete constraint — a latency target, a traffic shape, a cost ceiling — and works through the options until one of them is clearly the right trade-off. You will find serverless compute and its cold starts, scale-to-zero as a deliberate budget strategy, media delivery through CDNs, event-driven designs, and the hidden fixed costs that make a “free tier” architecture surprisingly expensive.
Where numbers appear, they are either measured or clearly labelled as illustrative, and pricing always links to the official page so you can check it against today’s rates.
Featured
A design review of a Google Cloud image pipeline: signed uploads, fixed variants, immutable names and a CDN, with failure and deletion paths worked through.
15 min
· 12 min
Batch work behaves badly behind an HTTP endpoint. Compare Cloud Run jobs and services on retries, parallelism and exit codes, and shard work across tasks.
12 min
· 10 min
Start with a budget that emails a monitored channel, then add a Pub/Sub relay, usage alerts and a heartbeat that notices silence when that is worth building.
10 min
· 8 min
A static bundle in Cloud Storage is cheap to run and easy to get wrong: which Cache-Control values to set, how to upload them, and how Cloud CDN behaves.
8 min
· 9 min
Turn the Cloud Billing export into five queries that explain a Google Cloud invoice: what is fixed, what scales, what is a credit, and who owns each line.
9 min
· 11 min
Pick the platform by workload shape, not habit: when Cloud Run services, Cloud Run functions, jobs or GKE Autopilot fit, and how to move between them later.
11 min
· 16 min
An erasure request touches your database, search index, buckets, logs, analytics, backups and vendors. Here is how I make that deletion retryable and provable.
16 min
· 15 min
A scheduled sync pays to refresh every record, including the ones nobody reads. Refreshing stale data when someone looks makes cost follow attention.
15 min
· 15 min
Measure which phase of a Cloud Run cold start is slow, match it to the lever that shortens it, and pay for warm instances only where the latency earns it.
15 min
· 15 min
A design review of a Google Cloud image pipeline: signed uploads, fixed variants, immutable names and a CDN, with failure and deletion paths worked through.
15 min